Vannarot Roeung
Recent posts by Vannarot Roeung
Google Flow Merges Whisk, ImageFX, and Video into One AI Studio
Google merged three separate AI creative tools into one platform in March 2026. Flow now combines Whisk (mood boards and collages), ImageFX (text-to-image generation), and Flow's original video capabilities into a single workspace powered by Veo 3.1 and Nano Banana
Kiwi-Edit: Open-Source AI Video Editing with Text and Reference Images
Kiwi-Edit, a new open-source video editing framework from NUS ShowLab, launched on March 5, 2026, combining text instruction guidance with reference image control to handle both global and local video edits at 720p resolution
Luma Creative AI Agents Launch With Uni-1 Model
Luma launched Creative AI Agents on March 5, 2026, powered by its new Uni-1 model. The system handles end-to-end creative production across text, image, video, and audio, coordinating with external models like Google Veo 3, Nano Banana Pro, ByteDance Seedream, and ElevenLabs
GPT-5.4 Launches with Native Computer Use Mode
OpenAI launched GPT-5.4 on March 5, 2026, introducing native Computer Use mode that lets the model read screens and control mouse and keyboard inputs directly. The release also brings a 1M token context window, 33% fewer factual errors across benchmarks, and 47% token efficiency gains
Google Flow Merges Whisk, ImageFX, and Video Into One AI Creative Platform
Google just killed three products and replaced them with one. On February 25, 2026, the company merged Flow, Whisk, and ImageFX into a single creative workspace powered by Veo 3.1, Nano Banana, and Gemini
OpenAI Codex App Launches on Windows with Free Tier Access
OpenAI released the Codex coding agent app for Windows on March 4, 2026, expanding beyond the Mac-only launch. The app is temporarily free for ChatGPT Free and Go subscribers, with doubled rate limits for all paid tiers
Helios: Open-Source 14B Video Model Hits Real-Time Speed
Helios is a 14-billion-parameter open-source video model that runs at 19.5 FPS on a single NVIDIA H100, generating up to 60 seconds of video in real time. Built by Peking University, ByteDance, and Canva, it launched under the Apache 2.0 license on March 4, 2026
GPT-5.3 Instant Drops the Cringe and Cuts Hallucinations by 27%
OpenAI released GPT-5.3 Instant on March 3, 2026, replacing GPT-5.2 Instant as the default model in ChatGPT. The update reduces hallucinations by 26.8% on high-stakes queries, eliminates the preachy tone users complained about, and ships to all ChatGPT users including the free tier
FireRed-Image-Edit 1.1: Xiaohongshu SOTA Open-Source Image Editor
Xiaohongshu's Intelligent Creation team released FireRed-Image-Edit 1.1 on March 3, 2026, achieving a composite score of 7.94 across five authoritative benchmarks and claiming the top spot among open-source image editing models
Qwen 3.5 Small: Edge AI Models Run on Phones
Alibaba's Qwen team just released Qwen 3.5 Small, a series of four open-source models ranging from 0.8B to 9B parameters that run on phones, IoT devices, and browser tabs. The smallest model operates on a seven-year-old Samsung Galaxy S10E at 12 tokens per second
AMD Ryzen AI 400 Brings AI Desktop CPUs and ROCm to Consumers
AMD launched the Ryzen AI 400 Series desktop processors at MWC 2026 on March 2, marking the first time its XDNA 2 NPU and RDNA 3.5 integrated graphics are available in a desktop form factor
OpenAI Raises $110 Billion at $730 Billion Valuation
OpenAI closed a $110 billion funding round on February 27, 2026, the largest private financing in history. The deal values the company at $730 billion, more than doubling its $300 billion valuation from just 12 months earlier
Meta Signs $50M-Per-Year AI Content Deal with News Corp
Meta Platforms signed a three-year AI content licensing deal with News Corp worth up to $50 million per year. The agreement gives Meta access to News Corp's US and UK news archives to train AI models and power real-time responses in Meta AI products across Facebook, Instagram, WhatsApp, and Messe...
Google Gemini 3.1 Pro Doubles Reasoning Performance
Google released Gemini 3.1 Pro on February 19, 2026, reporting more than double the reasoning performance of its predecessor on key benchmarks. The model scored 77.1% on ARC-AGI-2, a test designed to measure general intelligence capabilities
GLM-5: China Ships a 744B Open-Source AI Model Trained on Huawei Chips
Zhipu AI released GLM-5 on February 13, 2026, a 744 billion parameter open-source model trained entirely on Huawei Ascend chips without a single NVIDIA GPU
Apple Confirms Gemini-Powered Siri Overhaul for iOS 26.4
Live test 17:52:00
Apple Rebuilds Siri With Gemini AI: What the $1B Deal Means for Creators
Apple just admitted it cannot build a competitive voice assistant alone. On February 12, 2026, the company announced a multi-year partnership with Google to rebuild Siri from the ground up using Gemini AI
Suno V5 Studio Turns AI Music Into a Real DAW
Suno released major updates to V5 Studio on February 6, 2026, adding warp markers, FX removal, alternate takes, and time signature control
NASA Perseverance Completes First AI-Planned Drive on Mars Using Claude
NASA's Jet Propulsion Laboratory successfully completed the first autonomous drive on Mars where the rover's path was planned using Anthropic's Claude AI. Perseverance completed two separate drives totaling nearly 1,500 feet, with the first drive covering 689 feet and the second spanning 807 feet
Big Tech Commits $650 Billion to AI Infrastructure in 2026
The four largest technology companies have collectively committed $650 billion toward AI infrastructure spending in 2026, marking a dramatic 67 percent increase from 2025 investments
Perplexity Launches Model Council: Run GPT, Claude, and Gemini in Parallel
Perplexity launched Model Council on February 5, 2026, a feature that runs the same query across multiple frontier AI models simultaneously and synthesizes a single, cross-validated answer. The feature is available to Perplexity Max subscribers at $200 per month
OpenAI Launches GPT-5.3-Codex: The Model That Built Itself
On February 5, 2026, OpenAI shipped GPT-5.3-Codex within minutes of Anthropic's Claude Opus 4.6 release. This wasn't coincidence
Claude Opus 4.6: Agent Teams, 1M Token Context, and What It Means for Creators
Anthropic released Claude Opus 4.6 on February 5, 2026, introducing agent teams that coordinate across parallel tasks, a 1 million token context window (5x the previous limit), and 128K output tokens
Google Gemini Hits 750 Million Monthly Active Users
Google announced that Gemini has reached 750 million monthly active users, representing growth of 100 million users from the previous quarter's 650 million